Self-Hosted AI: The Infrastructure Reality Underneath the Hype

Tensor9
Tensor9

Share this:

The AI hype cycle focuses on models and benchmarks. But for enterprises deploying AI workloads, the real challenge is infrastructure: who controls the environment, who gets access when something breaks, and how to avoid turning the snowflake problem into a blizzard.

In this panel conversation, Nicky Pike (Field CTO, Coder) and Michael Ten-Pow (CEO, Tensor9) dig into what they’re seeing on the ground: why organizations are pulling AI workloads back from cloud APIs, why self-hosted deployment is a governance problem dressed up as an engineering problem, and what the rise of agentic SREs means for the future of on-prem software.